CodeTurn for Automated Mainframe Code Conversion

Mainframe applications often hold decades of business rules that companies still rely on for billing, payments, and other critical operations. Amdocs CodeTurn uses configurable rules to convert supported mainframe code to Java, C#, or another COBOL dialect. Teams use it when they want to preserve established business behavior while changing the language, compiler or execution environment.

Consider modernizing when routine changes take too long, critical application knowledge sits with a few specialists, or existing code is hard to integrate with new services. Automated code conversion is especially useful when the same legacy coding patterns recur across many programs. With CodeTurn, engineers can configure a rule for a shared pattern, apply it consistently across the affected programs, and rerun the conversion and tests as they refine the rule.

This article explains what CodeTurn can automate and how a representative conversion helps teams evaluate the generated code and verify that it produces the expected business results.

Key Takeaways

• CodeTurn applies configurable conversion rules across an application portfolio.
• DataTurn handles supported data conversion, while TestMatch and DataMatch help verify application and data behavior.
• A sample conversion lets teams inspect generated code before committing to a migration wave. 

What CodeTurn converts

A COBOL program rarely works alone. It can use shared copybooks, call other programs and contain statements that access CICS or databases such as Db2, IDMS or Adabas. The first step is to inventory those dependencies and confirm which combinations the conversion supports. Job control, external interfaces and third-party products may require complementary tooling or separate engineering.

CodeTurn uses parsers, analyzers and generators to apply rules tailored to the source and target. If several programs share an unsupported construct, engineers can resolve the pattern centrally and rerun the affected code. Configuration can also reflect target coding standards, so the team can review the result against its own expectations.

How CodeTurn fits into Amdocs modernization

CodeTurn handles the repeatable transformation step. Amdocs can use agentic AI to help explain undocumented logic, surface dependencies and plan changes. Engineers review those findings, govern conversion rules and use test evidence to decide what is ready. CodeTurn can work alongside DataTurn when a program’s data access has to change with its language.

What a demo can tell you

Contact our experts for a demo and use a sample code to examine naming, program structure, runtime dependencies and handling of difficult language features. Include at least one example with shared copybooks or data access. Record what was included so a reviewer can tell which findings apply to the broader estate.

For a production decision, add end-to-end tests that cover online transactions, overnight jobs, database changes and failure recovery. Ask the team that will own the application to trace a rule, debug a failure and make a small change. Their experience is an early signal of how maintainable the converted estate will be.